A significant barrier to effective research, diagnosis, and treatment of stuttering has been a long history of viewing it as a simple unitary disorder in spite of its immense diversity in presentation and responsiveness to treatment. The broad objective of the first 5-year cycle (plus no cost extension) of this project was to study stuttering onset and development by positing that subtypes account for the diversity. Our theoretical orientation rests on a model that stuttering involves complex, multilevel and dynamic factors that interact to ultimately disrupt speech production. We view stuttering as a complex developmental disorder encompassing subtypes that are both genetically and environmentally shaped, and expressed through differences in the language, motor, and temperament domains, allowing us to parcel out variability that is predictive of recovery, chronicity, and therapeutic efficacy. Specifically, our project has been testing the hypothesis of distinct subtypes in developmental stuttering with data from over 200 longitudinal and cross-sectional participants in three age groups, combined with data from control subjects as well as from children's parents (a total of 851 participants). The project has a three-part agenda: Part 1: The core longitudinal study of early stuttering maps the longitudinal pathways of stuttering development in preschool children across the domains of motor skills, language skills, and temperament/personality in relation to fundamental epidemiologic and genetic factors. Part 2: The cross-sectional studies of school-age children and adults extend the subtype hypothesis to older individuals who stutter. Performance is measured across the same domains as in Part 1. Part 3: The longitudinal/cross-sectional comparison integrates patterns in the longitudinal study with those in school-age children and adults. The combination of longitudinal with cross-sectional data thus allows for the first integration of multidimensional traits associated with stuttering across the age span from preschool to adulthood. The longitudinal data and much of the cross-sectional data have been collected during the first cycle of the grant and are ready for analyses aimed at subtypes identification. Because of the ambitious size of this multi- center project, and particularly in view of its longitudinal nature of the study which only yields critical data on persistency vs. recovery at its end, a good portion of the relevant data became available only recently, and can be now reduced and analyzed to allow the subtyping hypotheses to be definitively tested. This renewal proposal of the grant focuses on, first, profiling multi-domain attributes to identify subtypes of stuttering in preschool children; and second, extending the subtype hypotheses to older children and adults who stutter, by comparing profiles in the same domains in these populations.